the ultrastructural localization of the isozymes of … · during fixation. in all formaldehyde-...

THE ULTRASTRUCTURAL LOCALIZATION OF

THE ISOZYMES OF ASPARTATEAMINOTRANSFERASE IN MURINE TISSUES

84

J . M . PAPADIMITRIOU and'P . VAN DUIJN

From the Department of Pathology, Histochemical Section and Laboratory of Electron Microscopy,Leiden, The Netherlands

ABSTRACT

Two isozymes of aspartate aminotransferase have been demonstrated biochemically. Oneisozyme is found in the mitochondrial fraction of the cytoplasm, the other ("soluble") inthe supernatant. Both isozymes can be demonstrated by the cytochemical technique of Leeand Torack, as reported in the preceding report . Aldehyde fixation rapidly inactivatesboth isozymes, especially the soluble one. Inactivation can be delayed by addition of keto-glutarate to the fixative . The ketoglutarate probably competes with the fixative for theactive site of the enzyme, thus protecting that region of the molecule . This enables adequatetissue preservation with enough remaining enzymatic activity to be demonstrated by theprecipitation of oxaloacetate as the lead salt from a medium containing a-ketoglutaric acidaspartic acid, and lead nitrate . Electron-opaque material was found not only in mito-chondria but, as the result of substrate protection, on the plasma membranes of many cellsincluding erythrocytes and bacteria, the limiting membrane of peroxisomes, and the trans-verse tubular system of striated muscle . Occasional centrioles, neurotubules, tubules in thetails of spermatozoa, the A-I band junction in myofibrils of striated muscle, and the groundsubstance between cisternae of endoplasmic reticulum in intestinal goblet cells also showedprecipitate. In all cases, replacement of L-aspartic acid by n-aspartic acid in the mediumresulted in unstained sections . The sensitivity of extramitochondrial sites to fixation, theneed of ketoglutarate as an agent for protecting the enzymatic activity during the fixationprocess, and the known presence of only soluble isozyme in erythrocytes indicate thatenzymatic activity at these sites can be attributed to the soluble isozyme . Localization ofthe soluble isozyme on the plasma membrane may be related to possible involvement indepolarization phenomena, amino acid transport, or synthesis of plasma membrane-boundmucopolysaccharides .

INTRODUCTION

In the preceding paper (29), it was establishedthat, in a model system containing the cytoplasmicfractions of rat liver, both the mitochondrial andthe "soluble" isozymes of aspartate aminotransfer-ase (E.C . 2 .6 .1 .1 ., glutamate-oxalacetate transami-nase) (2, 3, 5, 6, 7, 9, 22, 24, 25, 31) can be demon-strated by the technique recommended by Lee and

Torack (17, 19, 20). The rapid inactivation of thesoluble enzyme by aldehyde fixation was dimin-ished by the addition of ketoglutarate to the fixa-tive (29) . The ketoglutarate is thought to competesuccessfully with the fixative for the active site ofthe enzyme and thus delay inactivation. In addi-tion, the optimum substrate concentration for de-

THE JOURNAL OF CELL BIOLOGY • VOLCME 47, 1970 . pages 84-98

Dow

nloaded from http://rupress.org/jcb/article-pdf/47/1/84/1385710/84.pdf by guest on 02 July 2021

testing soluble enzyme activity was determined .The information obtained from the model systemstudies was applied to an ultrastructural investi-gation of the sites of activity of the aspartateaminotransferase isoenzyme in murine tissues .These findings are reported in the present report .

Earlier ultrastructural studies by Lee andTorack on the sites of aspartate aminotransferaseactivity had shown what was primarily the locali-zation of the mitochondrial isozyme (18, 19, 20)in both rat heart and rat liver . Extramitochondrialactivity was noted in hepatocytic peroxisomes(19), nuclear membranes (18, 19), and vesicularcomponents of cardiac muscle (18) . As regards thereaction product surrounding peroxisomes, Leeand Torack (19) suggested that it might be due tothe activity of the soluble isozyme . However, theseextramitochondrial sites were not always consist-ently demonstrated . With the techniques out-lined above, electron-opaque material, attributedto the activity of the soluble isozyme of aspartateaminotransferase, was unexpectedly found on theplasma membrane of many cell types, as well as onother cytoplasmic structure . It appears that thisisozyme, which biochemically is found in thesupernatant fraction of the cytoplasm, is in closeassociation with structural cytoplasmicponents .

MATERIALS AND METHODS

com-

Murine liver, heart, diaphragm, cerebral cortex, in-testine, and testis were selected for study . Adult al-bino mice were anesthetized, the thorax was opened,and the animal was perfused through the left ven-tricle for 2 min with either cold 1 % glutaraldehyde(Fluka AG, Buchs, Switzerland) or 3 .7% formalde-hyde (E. Merck AG, Darmstadt, Germany) in 0 .05M imidazole (Koch-Light Laboratories Ltd ., Coln-brook, Buckinghamshire, England) buffer (pH 7 .2)containing I g of sucrose/100 ml . Small pieces ofthe selected tissues (0.5 mm3) were then placed infresh fixative for a total of 7 or 12 min of fixation . Ina second group of animals, the same procedure wasfollowed but in this instance a-ketoglutarate was in-troduced in the fixative at a concentration of 117 mg /100 ml (8 mm) .

After fixation, tissues were washed for 1 hr in cold0.25 M sucrose solutions buffered with 0.05 M imida-zole (pH 7 .2) and then placed for 20 min in an incu-bation medium containing 4 mm of a-ketoglutarate(Koch-Light Lab ., England), 20 mm of L-aspartate(Koch-Light Lab ., England) and 6 mm of lead nitrate(pH 7 .5) . Incubation was performed at room tem-perature (20°-22°C) . After incubation, specimens

were washed for 5 min in 20 DIM of L-aspartate in0 .05 M imidazole (pH 7 .3) as recommended by Leeand Torack (19) . They were then postfixed in 1%osmium tetroxide dissolved in s-collidine buffer (pH7.3) for 1 hr, dehydrated in graded solutions ofethanol containing imidazole buffer (pH 7 .3), andembedded in Epon . Sections were cut on a Reichertultramicrotome and examined in a Siemens Elmiskopat an accelerating voltage of 80 kv. Some sectionswere stained with lead hydroxide; others were ex-amined unstained .

An incubation medium prepared as outlined abovebut containing D-aspartic acid instead of L-asparticacid provided for a series of controls .

For light microscopy, frozen sections of liver andmuscle were cut at a thickness of 10-15 µ, fixed for 2min in cold 3 .7% formaldehyde, and incubated inthe medium described above . After incubation, theywere treated for 5 min with a 2% sodium sulfidesolution, washed, and mounted in "Aquamount"(Edward Gurr Ltd ., London S .W. 14, England) .

RESULTS

Generally, tissues fixed in 170 glutaraldehydecontaining ketoglutarate regularly exhibited evi-dence of enzymatic activity, whereas the tissuesfixed in glutaraldehyde from which ketoglutaratewas omitted gave inconsistent demonstration of thereaction product, especially in extramitochondrialsites of enzymatic activity . An acceptable degreeof ultrastructural preservation was attained in mostcases, despite the rather short fixation periodsemployed.

Fixation with formaldehyde gave similar results .Again, preservation of enzymatic activity wasgreater when ketoglutarate was included in thefixative. Moreover, the combination of formalde-hyde and ketoglutarate preserved more sites ofenzymatic activity than the combination of 107cglutaraldehyde and ketoglutarate. Unfortunately,the ultrastructure was generally poorly preservedwhen formaldehyde was used, especially when thefixation times were short .

Liver

LIGHT MICROSCOPY : Brown reaction prod-uct (lead sulfide) was visible in all hepatocytes.Deposition of the reaction product was maximalin the periportal areas of the liver lobule (Fig . 1)and least in hepatocytes surrounding the centralvein .ELECTRON MICROSCOPY : Reaction product

was easily recognized by its electron opacity andwas seen to localize in the majority of mitochondria

PAPADIMITRIOU AND VAN DUIJN Aspartate Aminotransferase in Murine Tissues

85

Dow


within hepatocytes and sinusoidal cells (Fig . 2) .Some mitochondria, especially in tissues fixedwithout substrate protection, failed to show thepresence of enzymatic activity although adjoiningmitochondria within the same cell possessed dis-tinct deposits of reaction product (Fig . 2) . Thesedeposits were confined to the region between theinner and outer mitochondrial membranes andextended into the cristae mitochondriales (Figs .2, 3) Reaction product was not found within theinner mitochondrial chamber .

The limiting membrane of hepatocytic peroxi-somes also possessed electron-opaque deposits ofreaction product (Fig . 3) . These deposits were mostconspicuous in specimens fixed in the presence ofa-ketoglutarate, although some were found inspecimens fixed in its absence . Neither the matrixnor the nucleoid of peroxisomes showed evidenceof enzymatic activity .

When ketoglutarate was included in the fixative,deposits of reaction product became obvious overthe membrane lining the hepatocytic microvilli inthe spaces of Disse, in micropinocytic vesicles ofsinusoidal cells and occasionally parenchymal cells,and the plasma membrane of erythrocytes (Fig .3). These deposits were somewhat more con-spicuous when formaldehyde was substituted forglutaraldehyde in the fixation process . Occasion-ally deposits were found on the membrane liningthe bile canaliculi but not in any other regions ofthe plasma membranes lining the interhepatocyticspace .

Neither the nucleus nor the nuclear membraneof hepatocytes or sinusoidal cells possessed anytrace of reaction product . The endoplasmic re-ticulum and the Golgi complex also failed to showevidence of enzymatic activity .

Blood Vessels

The endothelial cells lining blood vessels in theheart, diaphragm, and brain all exhibited evidence

46

of enzymatic activity . Apart from its appearancein the scant mitochondria of endothelial cells, thereaction product was readily found within thecontents of most micropinocytic vesicles (Fig . 4) .Again, this was very evident in tissues fixed in thepresence of ketoglutarate . Endothelia of arterioles,capillaries, and venules all showed this type oflocalization . In addition, deposits were also identi-fied in some junction regions between adjoiningendothelial cells .

In tissues fixed for 12 min in 3 .7 % formaldehydeor for 7 min in 1 % glutaraldehyde, but alwayswith added ketoglutarate, deposits became obviousover the plasma membrane of endothelial cells(Fig . 5) . Erythrocytes in the lumen of these vesselsalso showed deposits on the cell membrane (Fig .5) . Pericytes and smooth muscle cells also exhibitedreaction product within micropinocytic vesiclesand over cell membranes .

Cardiac and Skeletal Muscle

Again, mitochondria between myofibrils orgrouped in the subsarcolemmal region showed thepresence of reaction product (Fig . 6) . The depositswere localized over the outer mitochondrialchamber and extended into the cristae mitochon-driales. In addition, enzymatic activity was mani-fest around transverse tubules (Fig . 6) but not inthe lumen of the tubule . These deposits were con-spicuous only when ketoglutarate was incorporatedduring fixation . In all formaldehyde- and a fewglutaraldehyde-fixed tissues, reaction product wasobvious over the sarcolemmal membranes (Fig . 7)and the plasma membranes of adjoining fibroblasts(Fig. 8), provided ketoglutarate was present inthe fixative . Direct connections between sarco-lemma and transverse tubules were not preservedin the preparations. It was thus not possible toexamine the enzymatic localizations at these sites .The region of the A-I band junction of myofibrilsalso exhibited dense lead deposits . Occasional

FIGURE 1 Section of mouse liver in which the dense deposits of lead sulfide indicate sites of activityof aspartate aminotransferase. Hepatocytes in the periportal area show more activity than those in thecentral region of the lobule . Formaldehyde fixation. Magnification, 700 X .

FIGURE 2 Liver fixed for 12 min in 1°%o glutaraldehyde . Electron-opaque deposits are present in theouter mitochondrial chamber and extend into the cristae mitochondriales . Some mitochondria (M) donot show evidence of activity . Perosixomes (arrow) also fail to demonstrate enzymatic activity . Leadhydroxide . Magnification, 24,000 X .

THE JOURNAL OF CELL BIOLOGY . VOLUME 47, 1970

Dow


87

Dow


centrioles in fibroblastic cells showed well defineddeposits within the lumina of their constituenttubular structures (Fig . 8) .

Nuclei and nuclear membranes remained un-stained . Neither the Golgi apparatus nor thecomponents of the sarcoplasmic reticulum showedenzymatic activity .

Brain

Cerebral tissues were difficult to preserve wellbecause the short fixation times used in order toavoid enzymatic inactivation failed to adequatelypreserve all morphological details . Nevertheless,blood vessels and neurones were moderately wellpreserved . As expected, mitochondria withinneurones, axons, and dendrites exhibited electron-opaque deposits in a distribution similar to thatfor other tissues . With short fixation times (7 min)with 1 % glutaraldehyde and ketoglutarate pro-tection, enzymatic activity was detected on thesurface of neuronal perikarya dendrites, theperiphery of occasional myelin sheaths, but not onthe limiting membrane of axons (Fig . 9) . Wheresynapses were preserved, the reaction product wasconfined to the gap region of the synaptic space(Fig . 10) . Occasional neurotubules showed leaddeposits (Fig. 11) . No other cellular componentshowed enzymatic activity .

Intestinal Columnar Epithelium

Mitochondrial enzymatic activity was obviousin the outer mitochondrial chamber . In tissuesprotected with ketoglutarate during fixation, re-action product was found along the periphery ofthe lining columnar cells and outlining the surfaceof the microvilli (Fig . 12) . In goblet cells fixed informaldehyde in the presence of ketoglutarate,enzymatic activity was detected in the small bandsof ground substance between cisternae of endo-plasmic reticulum in the paranuclear region, andsparsely over the mucoid secretions of the cell (Fig .16) .

88

Bacteria present in the gut lumen also showedthe presence of enzymatic activity on their surfaces(Fig . 12) . The dense lead deposits were localizednot on the capsule of the bacterial cell but mainlyon the cell membrane of the bacterium (Fig . 13) .

Testis

Apart from the mitochondria of a variety of celltypes, reaction product was present in the plasmamembrane of Sertoli cells and spermatozoa (Fig .17) but less frequently in the cell membranes ofspermatids and their precursors . In addition,electron-opaque deposits were seen in some cen-trioles and axial tubules in the tail of spermatozoa(Figs. 14, 15) . Substrate protection was necessaryfor the demonstration of all extramitochondrialsites of activity .

Controls

Tissues prepared in the same manner as indi-cated above but incubated in a medium containingD-aspartic acid instead of L-aspartic acid did notshow electron-opaque deposits at any site whereenzymatic activity was manifest in the experi-mental series (Fig . 18) . Specimens of liver, muscle,heart, brain, intestine, and testis were included inthese control experiments . Generally, the leadcontamination in these tissues was remarkably low .

DISCUSSION

There is little doubt that the addition of ketoglu-tarate to the fixative minimizes enzymatic inacti-vation and permits a more complete cytochemicaldemonstration of aspartate aminotransferase ac-tivity . These results are in accord with those ob-tained from model studies of the histochemicaltechniques used (29) . Such correlation calls forfurther monitoring of cytochemical reactions withmodel systems similar to that described, in theapplication of cytochemical methods to ultra-structural use . As expected from the model study,formaldehyde proved less detrimental for enzy-

FIGURE 3 Liver fixed for 12 min in 1% glutaraldehyde with added ketoglutarate . In addition to mito-chondria, peroxisomes (arrows), hepatocytic microvilli, micropinocytic vesicles (arrowhead), and theplasma membrane of erythrocytes show evidence of activity . Magnification, 20,000 X .

FIGURE 4 Arteriolar endothelium from striated muscle, fixed for 12 min in 1% glutaraldehyde andketoglutarate. Enzymatic activity is present in micropinocytic vesicles (arrows) and the junction region(arrowheads) . Lead hydroxide . Magnification, 28,000 X .

THE JOURNAL OF CELL BIOLOGY • VOLUME 47, 1970

Dow


89

Dow


matic activity than glutaraldehyde, although thelatter gives better fixation of ultrastructure . Acombination of both fixatives, as used by Lee andTorack (19), and substrate protection may proveideal in future electron microscopic studies.

The mitochondrial localization of aspartateaminotransferase activity was expected both froma number of biochemical studies (2, 3, 5, 6, 7, 9,22, 25, 31) and from the more recent ultrastruc-tural evidence in rat liver and heart (18, 19, 20) .This type of localization was readily obtained,and did not necessitate substrate protection for itsdemonstration within the time limits of the fixa-tion periods used .

The available evidence indicates that the mito-chondrial isozyme bears a close relationship to themitochondrial membranes since sonication anddetergents largely fail to solubilize the enzyme(1, 5) . Recently Schnaitman and Greenawalt (30)have reported that aspartate aminotransferase isintimately related to the inner membrane-matrixfraction of mitochondria . These results are in ac-cord with the finding of reaction product betweenthe inner and outer mitochondrial membranes andextending into the cristae mitochondriales, sug-gesting a close relationship to the inner mitochon-drial membrane . It was noted that, in somehepatocytes, not all mitochondria showed thepresence of enzymatic activity. It is not knownwhether this represents a fixation artifact or if it isan indication of a mitochondrial population withinhepatocytes which is relatively poor in aspartateaminotransferase . Light microscopical studiesindicate that in mouse liver the periportal hepato-cytes possess a greater degree of enzymatic activitythat the hepatocytes surrounding the central vein .This may be further evidence of a bimodal distri-bution of hepatocytic mitochondria . Swick et al .(33) have provided some evidence that this, in-deed, may be the case not only for aspartateaminotransferase but for other mitochondrialenzymes as well.

Biochemical studies have not hinted at any

90

particular localization for the soluble isozyme(2, 3, 5, 6, 7, 9, 22, 24, 25, 30) . The term "soluble"merely indicates that upon differential centrifuga-tion the isozyme is found in the supernatant frac-tion of the cellular homogenate, and it generallydoes not imply that the isozyme originates in anyparticular organelle . It was therefore somewhatunexpected that reaction product was present oncytoplasmic structures other than mitochondria,while similar structures in control tissues were notstained . These extramitochondrial sites becamethe necessary candidates for the localization ofsoluble isozyme activity in vivo . The effect of fixa-tion (by formaldehyde or glutaraldehyde) andsubstrate protection at these sites was noted sincethis had already been assessed for the soluble iso-zyme in experiments utilising the model system(29) . Thus the greater sensitivity of the solubleisozyme to fixatives and the greater beneficialeffect of substrate protection on this isozyme wereused as a guide to which type of isozyme was beinglocalized . Moreover, the enzymatic activity oferythrocytes, which have been shown to containonly the soluble isozyme (27), paralleled theactivity of the soluble isozymes in the model studies(29) .

All this indicates that the electron-opaque de-posits found in erythrocytes, bacteria, and plasmamembranes of other cells are due to the activity ofthe soluble enzyme. Though there is good evidencethat the deposits at these sites are due to the ac-tivity of the soluble enzyme, the question whetherthey demonstrate the exact site of the enzyme asit occurs in vivo is more difficult to answer .

During the fixation procedure the enzyme maymove from the cytoplasm or from some extracel-lular space to the membrane location . Whetherthis will be the case will depend on the velocity ofsuch diffusions, on the rate of the fixation process,and on the firmness with which the enzyme isbound to a certain structure . Independent evi-dence to exclude enzyme movements prior tofixation is difficult to obtain at the moment . The

FIGURE 5 Cerebral venule fixed in 3.7% formaldehyde and added ketoglutarate, for 19, min . Reactionproduct is present on the erythrocytic plasma membrane, the endothelial cell membrane (arrowhead),and the membrane of pericytes . Unstained. Magnification, 34,000 X

FIGURE 6 Striated muscle fixed in 1% glutaraldehyde and added ketoglutarate for 12 min . Reactionproduct is present in the outer chamber of mitochondria and surrounds transverse tubules (arrowheads) .Lead hydroxide. Magnification, 30,000 X .

THE JOURNAL OF CELL BIOLOGY • VOLUME 47, 1970

Dow


91

Dow


same holds true for the possibility that the pre-cipitated lead salt of oxaloacetic acid does not pre-cipitate exactly at the site of the enzyme but atsome distance from its original site which has somespecial affinity for lead or oxaloacetic acid.

The fact that the ultrastructural localization ofthe cationic isozyme is in accord with biochemicalfindings-which locate this enzyme in the mito-chondrial fraction-supports the suggestion thatthe cytochemical fixation and staining procedurealso has some merit in pointing to the cellularlocalization of the soluble enzyme . The fact thatbiochemical studies so far have not pointed to anyparticular localization of the soluble enzyme doesnot in itself prove that in the vivo localization ofthe enzyme is not in or on membranes . Loss ofcomponents originally situated on or in membranesduring the preparation procedures now in usecannot be excluded . In the case of sponge cells,it was found that macromolecular products local-ized at the cell surface could be washed off evenby treatment of the cells with cold seawater free ofdivalent cations (26) . Accepting the present cyto-chemical evidence to represent the true location,it may indicate that the enzyme is associated withthe glycocalyx of a number of cell types. Since theenzyme is present in the sera of normal animals(6, 7, 21, 31) one also has to consider whether theenzyme has precipitated onto the cell surface as aconsequence of the fixation procedure .

The presence of enzymatic activity within thedeeper endothelial micropinocytic vesicles indi-cates that, in this case, absorption of enzyme asthe result of fixation is unlikely . The presence of

92

enzymatic activity in the interendothelial junc-tions also indicates that at least part of the activitydetected may be inherent to a membranous orglycocalyceal component since the large molecularweight (24) (120,000) of the soluble isozyme wouldhinder its passage through the junction . Moreover,the presence of activity on the surface of cell typesnot directly in contact with the vascular space,such as bacteria in the lumen of the gut, favors tosome extent a primary plasma membrane locali-zation .

The physiological significance of plasmalemmallocalization remains conjectural . Glutamate, andto some degree aspartate, two substances involvedin the enzymatic catalysis of aspartate amino-transferase, induce cell membrane depolarization(4, 10, 11, 14, 16, 36, 37) . It is not known if theenzyme is involved in the selective transport ofamino acids across cell membranes, althoughglutamate will accumulate in isolated brain tissuesagainst a concentration gradient (32) .Cell membrane-bound soluble isozyme may

thus be implicated in glutamate-induced neuronaldepolarization (4, 10, 11, 16) . Biochemical evi-dence is available for its presence in the synapto-some fraction of brain (8) . Similarly, it may be ofimportance in the depolarization of striated muscleby glutamate as occurs in insect (14, 36, 37) andcrustacean muscle (34) . In this respect, its distri-bution near elements of the transverse tubularsystem which is involved in the inward spread ofdepolarization (13), and its resemblance to thedistribution of choline esterase (35), are intriguing .

The presence of reaction product on the limiting

FIGURE 7 Striated muscle fixed in 3.7% formaldehyde with added ketoglutarate for 12 min. Reactionproduct is present on the cell membrane (upper arrowheads) and on the A-I band junction (lower arrow-heads) . Unstained . Magnification, 90,500 X .

FIGURE 8 Fibroblast fixed in 3.7% formaldehyde with added ketoglutarate . Reaction product is pres-ent in the plasma membrane (arrowheads) and the component tubules of a subjacent centriole (arrow) .Unstained . Magnification, 70,000 X

FIGURE 9 Neurone (NE) fixed in 1% glutaraldehyde with added ketoglutarate for 7 min . Reactionproduct is present on the plasma membrane of the neurone (arrowheads) . Unstained . Magnification,35,000 X .

FIGURE 10 Mouse brain fixed for 7 min in 1% glutaraldehyde and added ketoglutarate . Reactionproduct (arrowheads) is present in the gap region of a synapse . Magnification, 60,000 X .

FIGURE 11 Mouse brain fixed for 7 min in 1% glutaraldehyde and added ketoglutarate . Reaction prod-uct is present in the neurotubules of a dendritic process (arrowheads) . Unstained. Magnification, 32,000 X


Dow


93

Dow


membranes of hepatocytic peroxisomes confirmsthe findings of Lee and Torack (19) in the rat .Enzymatic activity at this site is also more suscep-tible to the effects of glutaraldehyde fixation thanmitochondrial activity, again suggesting that suchlead deposits represent sites of activity of thesoluble isoenzyme of aspartate aminotransferase .A similar proposition has been made by Lee andTorack (19) . They suggest that the discrepancybetween these histochemical data and the bio-chemical findings of Baudhuin et al . (1), whofailed to demonstrate aspartate aminotransferasein peroxisomes, may be due to loss of this solubleisozyme during the fractionation procedure .

Generally, it seems that sites of activity of thebiochemically defined soluble isozyme are closelyrelated to structural and often membranous com-ponents of the cell . Perhaps its physico-chemicalrelationships to these structures are such that theenzyme molecules are easily detached during thetrauma of high speed centrifugation .Reaction product was detected in only a few

centrioles, neurotubules, and axial tubules of thetails of spermatozoa . Negative results may be dueto low enzyme concentrations, with subsequentfailure of formation of a reaction product (12) .The presence of enzymatic activity on all thesestructures and on cell membranes points perhapsto an underlying fundamental similarity . A uni-fying hypothesis based on biochemical and mor-phological similarities has been proposed by Mazia(23), in an attempt to link these diverse structures .The only situation in which enzymatic activitywas detected in the ground substance is in theintestinal goblet cells, and, even there, it appearedin close apposition to cisternae of endoplasmic

94 THE JOURNAL OF CELL BIOLOGY • VOLUME 47, 1970

reticulum . Since glutamate plays a role in theamination of amino sugars in the early stages ofmucopolysaccharide synthesis (15), aspartateaminotransferase may be required in increasedamounts in goblet cells which are active in theproduction of these materials .

These data suggest that the soluble isozyme ofaspartate aminotransferase may be involved inmany basic cellular mechanisms and reactionswhich as yet remain partly or completely unsolved .Further information regarding its biochemical andphysiological role should prove rewarding .

This work was performed while one of us (Dr . Papa-dimitriou) was in receipt of a C . J . Martin TravellingFellowship from the National Health and MedicalResearch Council of Australia .

We are grateful to Dr . W. Th. Daems for his helpand advice throughout the course of this work andin the preparation of the manuscript .Received for publication 2 February 1970, and in revisedform 10 April 1970.

REFERENCES

1 . BAUDHUIN, P ., H. BEAUFAY, Y. RAHMAN-LI, 0 .Z. SELLINGER, R . WATTIAUX, P . JACQUES, andC. DE DuvE. 1964 . Tissue fractionation studies .Intracellular distribution of monoamine oxi-dase, aspartate aminotransferase, alanine aminotransferase, D-amino oxidase and catalase inrat liver tissue . Biochem . J. 92:179 .

2. BORST, P ., and E. M. PESTERS. 1961 . The intra-cellular localization of glutamate-oxaloacetatetransaminase in heart . Biochim . Biophys. Acta .54:188 .

3. BOYD, J. W. 1961 . The intracellular distributionlatency and electrophoretic mobility of L

FIGURE 12 Mouse intestine fixed in 3.7% formaldehyde for 12 min with added ketoglutarate . Reactionproduct is seen on the microvilli of the columnar epithelial cells and surrounding bacterial cells in theintestinal lumen . Unstained . Magnification, 37,000 X .

FIGURE 13 Bacterial cell in intestinal lumen fixed for 12 min in 3.7% formaldehyde with added keto-glutarate . Reaction product is confined to the cell membrane (arrowheads) and not the capsule of theorganism. Unstained . Magnification, 55,000 X .

FIGURE 14 Mouse testis fixed for 12 min in 3.7% formaldehyde with added ketoglutarate . Reactionproduct is present on the cell membrane and a centriole of a developing spermatid . Unstained . Mag-nification, 75,000 X .

FIGURE 15 Mouse testis fixed for 12 min in 3.77 formaldehyde with added ketoglutarate . Reactionproduct is present on the cell membrane and the axial microtubules of spermatozoan tail . Unstained .Magnification, 52,000 X .

Dow


9 5

Dow


glutamate-oxaloacetate transaminase of ratliver . Biochem . J. 80:18.

4. CURTIS, D . R. 1965 . Action of drugs on singleneurones in the spinal cord and thalamus . Brit .Med. Bull . 21 :5 .

5. EICHEL, H. J ., and J . BUKOVSKY. 1961 . Intracel-lular distribution pattern of rat liver glutamic-oxaloacetic transaminase . Nature (London) . 191 :243.

6. FLEISHER, G. A ., C. S. POTTER, and K . G. WAKIM.1960. Separation of 2 glutamic oxalacetictransaminases by paper electrophoresis . Proc .Soc. Exp . Biol. Med. 103 :229 .

7. FLEISHER, G . A., and K . G. WAKIM . 1961 . Pres-ence of two glutamic oxalacetic transaminasesin serum of dogs following acute injury of theliver . Proc. Soc . Exp . Biol. Med . 106 :283 .

8. FONNUM, F. 1968 . The distribution of glutamatedecarboxylase and aspartate transaminase insubcellular fractions of rat and guinea pigbrain . Biochem. J. 106 :401 .

9. GAULL, G., and C. A. VILLEE . 1960. Effects ofsome dicarboxylic acids on succinate oxidationin vitro. Biochim . Biophys. Acta . 39:560.

10. GIBSoN, I. M., and H. MCILWAIN. 1965 . Con-tinuous recording of changes in membranepotential in mammalian cerebral tissues invitro ; recovery after depolarization by addedsubstances . J. Physiol. (London) . 176 :261 .

11. HILLMAN, H . H., and H. MCILWAIN. 1961. Mem-brane potentials in mammalian cerebral tis-sues in vitro ; dependence on ionic environment .J. Physiol . (London) . 157 :263 .

12. HOLT, S. J ., and D. G. O'SULLIVAN . 1958 .Studies in enzyme cytochemistry . I . Principlesof cytochemical staining methods . Proc . Roy .Soc . Ser. B . 148 :465 .

13. HUXLEY, A . F., and R . E. TAYLOR. 1958 . Localactivation of striated muscle fibers . J. Physiol.(London) . 144 :246 .

14. KERKUT, G . A ., A. SHAPIRA, and R . J . WALKER .1965. The effect of acetyl choline, glutamicacid and GABA on the contractions of the

96


perfused cockroach leg . Comp . Biochem . Physiol .16 :37 .

15. KORNFELD, S., R. KORNFELD, E. NEUFELD, andP. J. O'BRIEN . 1964 . The feedback control ofsugar nucleotide biosynthesis in liver . Proc . Nat .Acad. Sci. U.S.A . 52 :371 .

16. KRNJEVIC, K ., and J. W. PHILLIS . 1963 . Ionto-phoretic studies of neurones in the mammaliancerebral cortex . J . Physiol . (London) . 165 :274 .

17 . LEE, S. H. 1968 . Histochemical demonstrationof glutamic oxalacetic transaminase . Amer. J.Clin . Pathol. 49:568.

18. LEE, S . H. 1969 . Ultrastructural localization ofglutamic oxalacetic transaminase in cardiacmuscle and cardiac mitochondrial fraction ofthe rat . Histochemie . 19 :99 .

19 . LEE, S . H., and R. M. TORACK . 1968 . Electronmicroscope studies of glutamic oxalacetictransaminase in rat liver cell . J . Cell Biol . 39 :716.

20. LEE, S. H., and R. M. TORACK. 1968 . A bio-chemical and histochemical study of glutamicoxalacetic transaminase activity of rat hepaticmitochondria fixed in situ and in vitro . J. CellBiol . 39 :725 .

21. MASSARAT, S ., and N . LANG. 1965. Immunologicand Enzymkinetik der Glutamat-Oxalacetat-Transaminase (GOT) and der Glutamat-Pyruvat-Transaminase (GPT) . Klin . Wo-

chenschr . 43:602 .22. MAY, W., M. MIYAZAKI, and R . G. GRENELL .

1959. The distribution of glutamic-oxaloacetictransaminase in rat brain . J. Neurochem . 4 :269 .

23. MAZIA, D . 1969 . Quoted in New approach to cell

division offered . Sci . Res . (N.Y.) 4(11) :23 .

24. MORINO, Y ., M. ITOH, and H. WADA. 1963 .

Crystallisation of 2-oxoglutarate-I-aspartatetransaminases from mitochondrial and solublefractions of beef liver. Biochem. Biophys . Res .

Commun . 13:348 .25. MORINO, Y ., H. KAGAMIYAMA, and H . WADA .

1964. Immunochemical distinction between

FIGURE 16 Intestinal goblet cell fixed in 3.7% formaldehyde with added ketoglutarate . Reaction

product is present in the ground substance between cisternae of endoplasmic reticulum . Unstained . Mag-

nification, 14,500 X

FIGURE 17 Mouse testis fixed in 3.7% formaldehyde with added ketoglutarate . Reaction product isfound on the plasma membrane of a developing spermatid and on that of an adjacent sertoli cell (arrow-heads) . Magnification, 36,000 X .

FIGURE 18 Mouse liver fixed for 12 min in 1%o glutaraldehyde with added ketoglutarate and incu-bated in a medium containing D-aspartic acid . Reaction product cannot be found at any site. Lead hy-

droxide. Magnification, 24,000 X

Dow


9 7

Dow


glutamic oxaloacetic transaminases from thesoluble and mitochondrial fractions of mam-malian tissue . J. Biol . Chem . 239:PC 943 .

26 . MOSCONA, A. A. 1968 . Cell aggregation : prop-erties of specific cell-ligands and their role inthe formation of multicellular systems . Develop.Biol. 18 :250.

27 . NISSELBAUM, J. S ., and O . BODANSKY . 1965 .Glutamic oxaloacetic transaminase in reticulo-cytes and erythrocytes . Science (Washington) .149:195 .

28 . OPPENHEIMER, S. B ., M . EDIDIN, C. W . ORR, andS . ROSEMAN . 1969 . An L-glutamine require-ment for intercellular adhesion . Proc . Nat . Acad.Sci. U.S.A . 63 :1395 .

29 . PAPADIMITRIOU, J. M ., and P . VAN DUIJN . 1970 .Effect of fixation and substrate protection onthe isoenzymes of aspartate aminotransferasestudied in a quantitative cytochemical modelsystem J. Cell Biol . 47 :71 .

30 . SCHNAITMAN, C., and J. W . GREENAWALT . 1968 .Enzymatic properties of the inner and outermembranes of rat liver mitochondria . J. CellBiol . 38 :158 .

31 . SCHMIDT, E., W . SCHMIDT, and P. OTTO . 1967 .

98


Isoenzymes of malic dehydrogenase, glutamicoxaloacetic transaminase and lactic dehydro-genase in serum in diseases of the liver . Clin .Chim . Acta . 15 :283 .

32 . STERN, J. R., L. V . EGGLESTON, R. HEMS, andH . A. KREBS . 1949 . Accumulation of glu-tamic acid in isolated brain tissue . Biochem . J.44:410.

33 . SWICK, R. W ., J . L . STRANGE, and J . F . THOMSON .1967 . The retrogeneous distribution of mito-chondrial enzymes in normal rat liver . Biochem-istry . 6 :737 .

34 . TAKEUCHI, A., and N . TAKEUCHI . 1964. Theeffect on crayfish muscle of iontophoreticallyapplied glutamate . J . Physiol. (London) . 170 :296 .

35 . TERAVAINEN, H. 1969 . Distribution of acetyl-cholinesterase in extraocular muscle fibres ofthe rat . Histochemie. 18 :174 .

USHERWOOD, P. N . R ., and H . GRVNDFEST . 1965 .Peripheral inhibitions in skeletal muscle ofinsects . J. Neurophysiol. 28:497.

USHERWOOD, P. N. R ., and P . MACHILI . 1966 .Chemical transmission at the insect excitatoryneuromuscular synapse . Nature (London) . 210 :634 .

36.

37 .

Dow


the ultrastructural localization of the isozymes of … · during fixation. in all formaldehyde-...

Documents